Confusing economic statistics in the time of COVID-19


The economic statistics that aid us in understanding the current state of the economy are in the midst of an unusual, if not unprecedented, upheaval.

Combined with the equally unparalleled oscillations of the economy through the early months of COVID, economy watchers are naturally confused. Add to this the tendency toward dissembling that accompanies elections, and we are poised for a couple months of economic confusion. This column is aimed at relieving some of that confusion.

The staff of the U.S. Bureau of Labor Statistics, Bureau of Economic Analysis and Census survey teams are responsible for compiling and reporting data on employment, GDP and other economic data. They are, by far, the best economic statistical services ever devised. Governments have been at this for more than a millennium, and the work done today is first-rate and thus far immune from political shenanigans. Americans should be pleased with the economists, statisticians and data scientists doing this work. They are not perfect.

The survey tools that give us the monthly unemployment rate were designed for personal contacts. COVID naturally interrupted that, so the response rate of phone calls plummeted. This introduced a great deal of uncertainty about the answers given to survey questions. This survey reports that roughly 11.2 million fewer people are working today than in January, of which roughly 4 million left the labor force. This yields in August an unemployment rate of 8.2%.

We also have an alternative, administrative measure of unemployment. This is the number of applications for unemployment insurance. In the very same week the survey told us we had 11.2 million people out of work, the administrative data told a startlingly different story. There were 13.79 million unemployed in the regular state system, another 13.5 million on pandemic unemployment assistance, and just under 2 million more in other programs. This is a total of 29 million unemployed and receiving benefits.

These data put our national unemployment rate at roughly 18.1% in August. Even accounting for a 15% error rate due to fraud or misreporting, the real U.S. unemployment rate is nearly 16%. The error rate is not that high; I simply want to be conservative and acknowledge the potential for errors in overcounting the unemployed.

If we could measure it effectively, the actual U.S. unemployment rate hasn’t been this high since before World War II. Given the ongoing effects of COVID, we won’t sort this out until next year, if then. But, by my reckoning, the actual unemployment rate is between 16% and 18%.

So, as we march toward the election, Americans can be certain that the official statistics offer very little truth about the level of employment. Again, that is not due to perfidy among federal statisticians. On the contrary, they are doing high-quality work with the utmost integrity. But, until we are released from COVID’s grip and can look retrospectively at the data, we must simply adjust to living in uncertain times. But uncertainty isn’t our only problem.

Misunderstanding the way we report changes to the economy can also easily lead to confusion. In the second quarter, our economy shrank by 9.1%, or at a compounded annualized rate of 37.1%. That is far and away the worst quarter on record. This quarter will be the best quarter of economic growth on record. We report GDP growth compared to the last quarter. Let’s say the economy grows at exactly the same rate it fell in the second quarter (9%). That wouldn’t erase the losses from second quarter.

If the economy shrinks by 9.1% in the second quarter, then recovers by 9.1% in the third quarter, the actual loss over two quarters will be roughly 5 percent. So, on the morning of Oct. 29, we will get some eye-popping news about economic growth, and economists will be quickly explaining that a 25% or 30% growth rate in third quarter really means the U.S. economy still shrank by more in six months than in any other time in history.

The third big challenge to understanding the state of the economy is the difference between permanent and part-time employment. Workers who’ve been temporarily laid off typically are recalled in a matter of months. Those who face permanent job losses take much longer to find work. The growing number of permanently unemployed is the long-term problem for the economy. Moreover, these numbers are surely undercounted, because they also come from the survey data. As of August, the number of permanently unemployed is growing at record pace.

The size of the problem is already enormous. At current levels, it will take perhaps six years to shrink these numbers back down to the January 2020 level. So, even if temporary layoffs return, we face a long, long slog to normalcy. So, as the election nears, be wary of claims about economic performance. There will be good news and bad news, but much of what you read or hear won’t be close to the truth.

Michael J. Hicks is the director of the Center for Business and Economic Research and the George and Frances Ball distinguished professor of economics in the Miller College of Business at Ball State University. His column appears in Indiana newspapers. Send comments to awoods@

No posts to display